CN104802793A - Method and device for classifying a behavior of a pedestrian when crossing a roadway of a vehicle as well as passenger protection system of a vehicle - Google Patents
Method and device for classifying a behavior of a pedestrian when crossing a roadway of a vehicle as well as passenger protection system of a vehicle Download PDFInfo
- Publication number
- CN104802793A CN104802793A CN201510030114.5A CN201510030114A CN104802793A CN 104802793 A CN104802793 A CN 104802793A CN 201510030114 A CN201510030114 A CN 201510030114A CN 104802793 A CN104802793 A CN 104802793A
- Authority
- CN
- China
- Prior art keywords
- pedestrian
- vehicle
- behavior
- environmental information
- asked
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/23—Recognition of whole body movements, e.g. for sport training
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30241—Trajectory
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30248—Vehicle exterior or interior
- G06T2207/30252—Vehicle exterior; Vicinity of vehicle
- G06T2207/30261—Obstacle
Abstract
A method for classifying a behavior, of a pedestrian when crossing a roadway of a vehicle, includes reading in a sensor signal to detect the pedestrian and at least one piece of surroundings information regarding surroundings of the pedestrian. The sensor signal represents here a signal of at least one sensor of the vehicle. The method further includes ascertaining at least one physical variable of a correlation between the pedestrian and the at least one piece of surroundings information. Finally, the method includes classifying the behavior of the pedestrian using the at least one physical variable.
Description
Technical field
The present invention relates to a kind of method, corresponding device, corresponding computer program, corresponding storage medium and passenger protection system for classifying to the behavior of pedestrian when crossing the moving traffic lane of vehicle.
Background technology
In order to realize initiatively pedestrian protection system, usually need the prediction of pedestrian movement, namely in the estimation of the possible dwell regions of future time adventurous pedestrian.Can determine whether start emergency braking or Emergency avoidance based on such estimation, to avoid upcoming accident.
Summary of the invention
Within this context, by scheme described here give chapter and verse main claim for the method and apparatus of classifying to the behavior of pedestrian when crossing the moving traffic lane of vehicle, corresponding computer program, corresponding storage medium and passenger protection system.Favourable configuration results from corresponding dependent claims and following explanation.
This programme realizes a kind of method for classifying to the behavior of pedestrian when crossing the moving traffic lane of vehicle, wherein, said method comprising the steps of:
Read sensor signal, so that at least one environmental information detecting described pedestrian and the environment about described pedestrian, wherein, described sensor signal represents the signal of at least one sensor of described vehicle;
Ask at least one physical parameter of the relation between described pedestrian and at least one environmental information described; And
When using at least one physical parameter described, the behavior of described pedestrian is classified.
Pedestrian can be interpreted as the personnel in the region of the moving traffic lane resting on vehicle.Behavior based on pedestrian can infer the possible motion of pedestrian in the region of moving traffic lane or sense of motion.That can be by the behavior understanding of pedestrian such as the attention of pedestrian or prepare cooperation degree.Such as moving traffic lane can be interpreted as the road of the one-lane or multi-lane travelled by vehicle.Especially vehicle can be interpreted as self-propelled vehicle, such as car, load truck or motor bike.The environment of pedestrian and pedestrian can detect by means of the sensor of vehicle.The environment of pedestrian be can be understood as moving traffic lane for pedestrian also or for vehicle driver visible or accessible region at least partly.At this, except moving traffic lane, object-also can belong to environment as the vehicle of the vehicle parked and traveling or other obstacles such as trees or central island.Term environment is not limited to moving traffic lane, but all objects and other elements, as line and kerb edge also can belong to environment.Feature that detect when using sensor, environment or element can be represented by environmental information.Environmental information can pass through electric signal or machine sensible data representation.Therefore, environmental information can by the device reprocessing be applicable to.Environment such as relates to the area, front of vehicle, and wherein, pedestrian can rest in the region in this area, front.Such as sensor can be interpreted as the pick up camera of the orientation of area forward of vehicle.Environmental information can be interpreted as the information about environment, described information is important relevant for the estimation of pedestrian behavior.Such as environmental information can represent obstacle, free surface or other objects, such as, other pedestrian in pedestrian's environment or other vehicle also or kerb edge with the trend of pavement marker.Be expert between human and environment information and can there is relation.Described relation can be interpreted as spatially, relation on temporal or space-time, it can be expressed by corresponding physical parameter.Such as physical parameter can be interpreted as speed, acceleration/accel, distance or the observed reading of time period.
This programme based on understanding be, the information from pedestrian's environment can be used for determining the possible behavior of pedestrian when traverses rows track.
The reliable prediction of possible pedestrian movement such as can be implemented based on the behavior determined like this.
The relevant prediction of such context such as also can be implemented when using and directly distributing to the feature of pedestrian.
These information or feature such as can be determined from the data of the video environment sensing mechanism of vehicle, directed especially forward three-dimensional video-frequency pick up camera.
Therefore, the utilization of prediction can be improved compared to conventional approach and especially reduce erroneous trigger rate when vehicle Emergency avoidance.
In order to avoid erroneous trigger, can apply a kind of method in emergency brake system, described method considers all possible track that pedestrian may advance naturally in prediction.At this, the only just trigger emergency braking when all possible track in described track or very large part such as 90% causes collision.
For being configured to by dodging manipulation, the system with the collision of pedestrian may being avoided in combination with automatic emergency brake, maximum hypothesis may be unaccommodated, because described maximum hypothesis may cause the too high estimation of dodging width of actual needs or not allow to dodge the determination of direction and width.Pedestrian may stop in the future potentially on multiple diverse location.Applicable dodging manipulation to still plan, the more conservative hypothesis about dwell regions in the future can be applied.Such as can suppose, pedestrian continues motion with constant speed.But this is not applicable in all cases.Such as pedestrian crosses road at him and not long ago often stopped, because this pedestrian has seen close vehicle.In order to consider this point, the stop line that pedestrian is possible can be determined.Such as motion feature can be utilized, such as, light stream in the leg and upper body area of pedestrian, to predict whether pedestrian stops or cross road for this reason.
According to a kind of embodiment of this programme, described method comprises behavior that basis is classified in the step of described classification, described pedestrian and determines the step of at least one possible track of described pedestrian.Possible track can be interpreted as the possible motion trend of pedestrian when traverses rows track.By determining that at least one possible track of pedestrian can identify the possible collision of pedestrian and vehicle ahead of time.
According to another embodiment of this programme, described method can comprise the step being provided for the activation signal of the personal protection means activating described vehicle according to described possible track.Such as personal protection means can be interpreted as device for passenger protection, such as, with the form of one or more safety air bag, also or be interpreted as pedestrian protection, such as, with the autobrake of vehicle or the form dodging manipulation.This embodiment by means of this programme can reduce the risk of injury for pedestrian, Vehicular occupant or other traffic participants.
In addition, in the step of described reading, the environmental model of the environment of described pedestrian can be created when using described sensor signal, to detect described pedestrian and described environmental information.Environmental model can be interpreted as the detailed data bank with pedestrian's environment can be described in real time.Can very accurately and reliably detect lines human and environment information when using such environmental model.
In addition, can classify to the behavior of pedestrian when using HMM and/or SVMs and/or fuzzy logic and/or neuroid in the step of classifying.The robust iterative of pedestrian behavior can be implemented thus.
In addition at least one possible stop line of pedestrian can be detected as environmental information in the step read.The speed of relative movement between pedestrian and possible stop line can be asked in the step asked at this.Can classify to the behavior of pedestrian when using speed of relative movement in the step of classifying.Possible stop line can be interpreted as in esse or virtual line, pedestrian stopped at this line place before his traverses rows track.Possible stop line such as can relate to carriageway marking or moving traffic lane edge.Can estimate pedestrian is whether or in which moment traverses rows track by means of the speed of relative movement between pedestrian and possible stop line.
According to another embodiment of this programme, the pedestrian at least one traverses rows track can be detected as environmental information in the step read.Spacing between the pedestrian that this can ask for pedestrian and described traverses rows track in the step asked for.Can classify to the behavior of pedestrian when using described spacing in the step of classifying.Can estimate by means of this spacing, whether pedestrian directly follows the pedestrian in traverses rows track.
In addition, can detect in the step read at least one of pedestrian possible cross position as environmental information.Pedestrian and the possible distance of crossing between position can be asked in the step asked at this.Can classify to the behavior of pedestrian when using described distance in the step of classifying.Such as possible position of crossing can be interpreted as public car station, building, zebra crossing or traffic signal lamp.Also can be determined the probability in pedestrian's traverses rows track with high precision by this embodiment of this programme.
Also can detect when using GPS information and/or cartographic information in the step detected according to another embodiment of this programme and cross position.High precision can be realized thus when asking for pedestrian and the possible distance of crossing between position.
In addition, the possible optical axis between vehicle and pedestrian can be detected as environmental information in the step read.The time gap of the possible optical axis can be asked in the step asked at this.Can classify to the behavior of pedestrian when using described time gap in the step of classifying.The possible optical axis can be interpreted as the free surface between vehicle and pedestrian.The possible optical axis can determine the relative visibility of pedestrian and vehicle driver.Therefore, time gap such time length can be interpreted as, vehicle can be seen when considering that the visual field hides this duration pedestrian.This time gap is larger, and the probability that vehicle is perceived by pedestrian is larger.This embodiment by means of this programme can assess the attention of pedestrian.
Body size and/or the line of vision of pedestrian can also be detected in the step read.At this, the behavior to pedestrian can also classify according to described body size and/or line of vision in the step of classification.Also the reliable assessment of pedestrian's attention can be realized by this embodiment of this programme.
The scheme proposed at this also realizes a kind of equipment, and described equipment is configured to the step of the modification implementing or realize the method proposed at this in corresponding device.By of the present invention with the enforcement flexible program of apparatus-form can solve fast and efficiently the present invention based on task.
Equipment can be understood as processes sensor signal at this and exports the electric equipment of control signal and/or data-signal accordingly.Described equipment can have by hardware mode and/or the interface by software mode structure.In the structure pressing hardware mode, interface can be such as the part comprising the least congenerous of described equipment of so-called system ASIC.But also possible that, interface is independent integrated circuit or is made up of discrete parts at least in part.In the structure pressing software mode, interface can be software module, and it such as coexists on a microcontroller with other software modules.
The computer program with program code is also favourable, described program code can be stored in machine sensible carrier, as in semiconductor memory, harddisk memory or optical memory and for when implementing according to the method for one of previously described embodiment during executive routine on the equipment corresponding to computing machine.
This programme also realizes a kind of machine sensible storage medium, and it has the computer program according to above-mentioned embodiment stored thereon.
Finally, this programme realizes a kind of passenger protection system of vehicle, and wherein, this passenger protection system has following characteristics:
For detecting at least one sensor of pedestrian and the environmental information about the environment of described pedestrian;
The device according to previously described a kind of embodiment be connected with at least one sensor described;
Be configured by the personal protection means that described device activates.
Accompanying drawing explanation
Below the scheme that this proposes exemplarily is illustrated in reference to the accompanying drawings further.Wherein:
Fig. 1: the schematic diagram with the vehicle of passenger protection system according to an embodiment of the invention;
Fig. 2 a, 2b: for the schematic diagram of stop line applied in method according to an embodiment of the invention;
Fig. 3: for the diagram of circuit of an embodiment of method of classifying to the behavior of pedestrian when crossing the moving traffic lane of vehicle;
Fig. 4: the system architecture of passenger protection system according to an embodiment of the invention; And
Fig. 5: for the block scheme of an embodiment of device of classifying to the behavior of pedestrian when crossing the moving traffic lane of vehicle.
In the following description of advantageous embodiment of the present invention for illustrate in different figures and act on the same or analogous Reference numeral of similar element application, wherein, no longer repeated description is carried out to these elements.
Detailed description of the invention
Fig. 1 illustrates the schematic diagram of the vehicle 100 with passenger protection system 105 according to an embodiment of the invention.Vehicle 100 is positioned on moving traffic lane 110.Pedestrian 115 is arranged in the area, front of vehicle 100.Pedestrian 115 is just thinking traverses rows track 110.Passenger protection system 105 comprises sensor 120, personal protection means 125 and control setup 130.Control setup 130 is connected with sensor 120 and personal protection means 125.Sensor 120 is arranged in the front area of vehicle 100 and area, front towards vehicle 100 is directed.
Sensor 120 is configured to the environment detecting pedestrian 115 and pedestrian 115.Control setup 130 is configured to the representative environment of receiving sensor 120 and the sensor signal of pedestrian 115 and asks for the physical parameter of the relation between environment and pedestrian 115 when using this sensor signal.In addition, control setup 130 is configured to classify according to the behavior of described physical parameter by pedestrian 115.Be such as the behavior of the pedestrian in traverses rows track 110 in FIG by the behaviour classification of pedestrian 115 by control setup 130.
Control setup 130 is additionally configured to determine pedestrian 115 at least one possible track 135 when traverses rows track 110 according to the behavior of pedestrian 115.At this, control setup 130 can be configured to send activation signal for activating personal protection means 125 to personal protection means 125 according to possible track 135.
Personal protection means 125 be such as configured in response to activation signal reception cause vehicle 100 dodge motion 140.
Therefore, vehicle 100 can dodge the pedestrian 115 in traverses rows track 110 in time, to avoid the collision between vehicle 100 and pedestrian 115.
Fig. 2 a, 2b illustrate the schematic diagram of the stop line 200 for applying in method according to an embodiment of the invention.
Fig. 2 a illustrates the moving traffic lane 110 with vehicle 100.Two other vehicles 205 are parked in the moving traffic lane edge of moving traffic lane 110.Other vehicle 205 is one after the other arranged.Pedestrian 115 moves towards stop line 200 in the space between described other vehicle 205.Do not relate to direct appreciiable in fig. 2 a and relate to virtual stop line 200.Stop line 200 is produced by the connection lead between the automobile 205 parked at this.
Vehicle 100 is configured to detect stop line 200 and pedestrian 115 and the physical parameter asking for the relation between stop line 200 and pedestrian 115.Such as vehicle 100 is configured to ask for speed of relative movement v between stop line 200 and pedestrian 115 and classifies about the possible mode of motion of pedestrian 115 when traverses rows track 110 to the behavior of pedestrian 115 according to speed v.
Be different from Fig. 2 a, the stop line 200 in Fig. 2 b is exemplarily corresponding to the lane markings 210 at moving traffic lane edge.Such as also the intermediate strap 215 of moving traffic lane 110 or other marks can be detected as stop line 200.
Additionally depict another pedestrian 220 in figure 2b.Be different from the pedestrian 115 being positioned at moving traffic lane edge, this another pedestrian 220 most of traverses rows track 110.According to one embodiment of present invention, vehicle 100 is configured to also detect another pedestrian 220 and the spacing d asked between pedestrian 115 and another pedestrian 220.The possible behavior of pedestrian 115 can be inferred according to the value of spacing d.As long as spacing d has relatively little value, such as, compared to the same curb-to-curb width detected, then can be such as the behavior of the pedestrian in traverses rows track 110 by the behaviour classification of pedestrian 115.
According to another embodiment of the present invention, vehicle 100 is configured to the possible optical axis between detection vehicle 100 and pedestrian 115 and asks for the time gap t of the possible optical axis.Alternatively or additionally, can classify to the behavior of pedestrian 115 according to time gap t.
Fig. 3 illustrates the diagram of circuit of an embodiment for the method 300 of classifying to the behavior of pedestrian when crossing the moving traffic lane of vehicle.Method 300 comprises read sensor signal to detect the step 305 of pedestrian and at least one environmental information about the environment of pedestrian, and wherein, sensor signal represents the signal of at least one sensor of vehicle.In addition, method 300 comprises the step 310 of the physical parameter of the relation asked between pedestrian and at least one environmental information.Finally, method 300 is included in the step 315 of classifying to the behavior of pedestrian when using at least one physical parameter described.
According to one embodiment of present invention, detect in the step 305 read at least one of pedestrian possible cross position as environmental information.At this, in the step 310 asked for, ask for pedestrian and possible cross distance between position as physical parameter.Finally, when using this distance, the behavior of pedestrian is classified in step 315.
Fig. 4 illustrates the system architecture 400 of passenger protection system according to an embodiment of the invention.First, carry out the pretreatment of sensing data in step 405, described sensing data is such as provided by the three-dimensional video-frequency sensor towards vehicle front area orientation.Disparity map or segregator is such as created in pretreatment.The environmental model of vehicle environmental is created in step 410 in response to pre-treatment step 405 ground.Environmental model can comprise such as take figure, list object or free surface describe.
When use pretreated sensing data in step 405 and/or create in step 410 environmental model implement feature in step 415 and determine.In this relation such as determining pedestrian and virtual stop line or other pedestrians also or determine the time length of mutual observability of vehicle and pedestrian.
Step 415 ground determined in response to feature realizes the step 420 of behaviour classification.Such as this by the behaviour classification of pedestrian for " stopping " relative to " crossing "; " attentively " relative to " inwholwe-hearted " or " cooperation " relative to " uncooperative ".
The behaviour classification that last basis is implemented at step 420 which realizes the step 425 of the prediction based on the behavior through classification.
---such as from the classification of the determination of the disparity map of three-dimensional video-frequency sensing mechanism or the gray value from video image---environmental model is estimated according to the embodiment shown in Fig. 1,2a, 2b and 4 of the present invention, based on environmental sensor 120 and corresponding original data processing.This environmental model comprises the information about obstacle, such as, with the form of obstructions chart, occupancy grid (" taking figure "), free surface or with the form of object.
The feature of the classification of the behavior being used for pedestrian 115 is extracted based on these data.The possible classification of pedestrian behavior is that such as " pedestrian 115 stops "---namely helps to avoid accident---compared to " pedestrian 115 is without cooperative taking action " compared to " pedestrian 115 does not see this vehicle 100 " or " pedestrian 115 cooperative takes action " compared to " pedestrian 115 keeps its sense of motion ", " pedestrian 115 has seen this vehicle 100 " relative to " pedestrian 115 crosses road 110 ", " pedestrian 115 changes its sense of motion ".
The classification based on these features of pedestrian behavior such as can be passed through HMM (HMM) (that is each hidden state is corresponding to a pedestrian behavior), SVMs (SVM), fuzzy logic or neuroid (NN) and realize.Multiple feature can be merged by these methods.Therefore the robust iterative of pedestrian behavior can be realized.
Finally, the prediction matched can be implemented based on identified pedestrian behavior.The prediction that such context of such as pedestrian 115 is relevant may be used for controlling initiatively pedestrian protection system 105.
Multiple feature is described below, and they can jointly for the classification of pedestrian behavior.Also a subset of described feature only can be used at this.
Key character for pedestrian movement's classification is the relative motion of pedestrian 115 for stop line 200 that may be virtual.This imagines based on drag: if traffic makes to cross at once become impossible, then want the pedestrian 115 crossing road 110 to imagine a line, this pedestrian stops at this line place.Such stop line 200 is such as embodied as moving traffic lane border, such as kerb edge, lane markings 210,215 or the border to a region, vehicle 100 is estimated to travel this region, and such as, a region between two vehicles parked 205, as shown in fig. 2 a.
Can use positive acceleration or the negative acceleration of pedestrian 115 as feature, it is necessary, so that pedestrian 115 still stopped before stop line 200.If acceleration/accel is very high, then this can be considered as pedestrian's state " cross road 110 " " having ignored this vehicle 100 " or " without cooperative take action " mark.
Another feature is the motion of other pedestrians 220.If pedestrian 220 has crossed road 110, then this can be used as the pedestrian 220 before the pedestrian 115 standing in road edge place follows and cross the mark of road 110 equally.The spacing d with the pedestrian 220 crossed can be used at this as feature.
The size of pedestrian 115 can be used as the mark of its behavior equally.Little pedestrian 115 may be children, its with the probability larger compared to large pedestrian 115 adopt state " cross road 110 ", " having ignored this vehicle 100 " or " without cooperative take action ", this large pedestrian may relate to adult.
Pedestrian 115 and this vehicle 100 are visible, namely do not hide from self visual field of pedestrian 115, then the probability of pedestrian 115 " stop ", " having seen this vehicle 100 " or " cooperative taking action " is higher.On the contrary, the mutual observability of short time corresponds respectively to other behavior state.
Alternatively or additionally; ask for by means of GPS or map datum, pedestrian 115 and special place---the such as next-door neighbour of public car station, school or kindergarten, improve state " cross road 110 ", the probability of " having ignored this vehicle 100 " or " without cooperative taking action ".In addition, the existence of the city motor bus stopped at halting point can be detected when public car station.
Zebra crossing, traffic signal lamp or traffic sign improve the probability of state " pedestrian 115 crosses road 110 ".Right in order to ask for alternative state clearly---as " pedestrian 115 has seen this vehicle 100 " relative to " pedestrian 115 does not see this vehicle 100 " and " pedestrian 115 cooperative takes action " relative to " pedestrian 115 without cooperative take action ", this feature such as can be compared with the other feature in described feature.
Alternatively, the line of vision of pedestrian 115 can be used, to determine whether pedestrian has seen close vehicle 100.Can also determine that pedestrian 115 will complete which target or this pedestrian whether its sense of motion of Planning Change by means of line of vision.
Fig. 5 illustrates the block scheme of an embodiment for the device 500 of classifying to the behavior of pedestrian when crossing the moving traffic lane of vehicle.Device 500 can be control setup 130 shown in Figure 1.Device 500---segregator also referred to as pedestrian movement---comprises for read sensor signal to detect pedestrian and the unit 505 about at least one environmental information of pedestrian's environment, wherein, sensor signal represents the signal of at least one sensor of vehicle.Unit 505 is connected with the unit 510 of at least one physical parameter for asking for the relation between pedestrian with at least one environmental information.Device 500 finally also comprises and to be connected with unit 510 and to be configured to when using at least one physical parameter the unit 515 that the behavior of pedestrian is classified.
Described only exemplarily selects with embodiment illustrated in the accompanying drawings.Different embodiments can intactly or about each feature combination with one another.An embodiment also can be supplemented by the feature of another embodiment.
In addition, can repeat and perform steps of a method in accordance with the invention with the order being different from described order.
If embodiment comprises the "and/or" relation between fisrt feature and second feature, then this is appreciated that as follows: described embodiment not only has fisrt feature according to an embodiment, and has second feature; And according to another embodiment or only there is fisrt feature, or only there is second feature.
Claims (13)
1. the method (300) for classifying to the behavior of pedestrian (115) when crossing moving traffic lane (110) of vehicle (100), wherein, described method (300) comprises the following steps:
Read (305) sensor signal, to detect at least one environmental information (200,220) of described pedestrian (115) and the environment about described pedestrian (115), wherein, described sensor signal represents the signal of at least one sensor (120) of described vehicle (100);
Ask at least one physical parameter (d, the t, v) of the relation between (310) described pedestrian (115) and at least one environmental information described (200,220); And
When use at least one physical parameter described (d, t, v) (315) are classified to the behavior of described pedestrian (115).
2. method according to claim 1 (300), described method has determines the step of at least one possible track (135) of described pedestrian (115) according in the middle behavior that classify, described pedestrian (115) of the step (315) of described classification.
3. method according to claim 2 (300), described method has the step of the activation signal being provided for activating the personal protection means (125) of described vehicle (100) according to described possible track (135).
4. the method (300) according to any one of the preceding claims, wherein, the environmental model of the environment of described pedestrian (115) is created when using described sensor signal, to detect described pedestrian (115) and described environmental information (200,220) in the step (305) of described reading.
5. the method (300) according to any one of the preceding claims, wherein, at least one possible stop line (200) of described pedestrian (115) is detected as environmental information in the step (305) of described reading, wherein, speed of relative movement (v) between described pedestrian (115) and described possible stop line (200) is asked for as physical parameter in the described step (310) asked for, wherein, when using described speed of relative movement (v), the behavior of described pedestrian (115) is classified in the step (315) of described classification.
6. the method (300) according to any one of the preceding claims, wherein, in the step (305) of described reading, detect at least one cross the pedestrian (220) of described moving traffic lane (110) as environmental information, wherein, spacing (d) between described pedestrian (115) and the described pedestrian (220) crossing described moving traffic lane (110) is asked for as physical parameter in the described step (310) asked for, wherein, when using described spacing (d), the behavior of described pedestrian (115) is classified in the step (315) of described classification.
7. the method (300) according to any one of the preceding claims, wherein, in the step (305) of described reading, detect that at least one of described pedestrian (115) is possible crosses position as environmental information, wherein, in the described step (310) asked for, ask for described pedestrian (115) and describedly possible cross distance between position as physical parameter, wherein, when using described distance, the behavior of described pedestrian (115) is classified in the step (315) of described classification.
8. the method (300) according to any one of the preceding claims, wherein, in the step (305) of described reading, the possible optical axis of detection between described vehicle (100) and described pedestrian (115) is as environmental information, wherein, the time gap (t) of the described possible optical axis is asked for as physical parameter in the described step (310) asked for, wherein, when using described time gap (t), the behavior of described pedestrian (115) is classified in the step (315) of described classification.
9. the method (300) according to any one of the preceding claims, wherein, body size and/or the line of vision of described pedestrian (115) is also detected in the step (305) of described reading, wherein, in the step (315) of described classification also according to described body size and/or described line of vision the behavior to described pedestrian (115) classify.
10. one kind be configured to implement method according to claim 1 (300) device (130,500) in steps.
11. 1 kinds of computer programs, its setting is for implementing the institute of method according to claim 1 (300) in steps.
12. 1 kinds of machine sensible storage mediums, it has computer program according to claim 11 stored thereon.
The passenger protection system (105) of 13. 1 kinds of vehicles (100), wherein, described passenger protection system (105) has following characteristics:
For detecting at least one sensor (120) of the environmental information (200,220) of pedestrian (115) and the environment about described pedestrian (115);
The device according to claim 10 (130,500) be connected with described at least one sensor (120);
Be configured by the personal protection means (125) that described device (130,500) activates.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| DE102014201159.0A DE102014201159A1 (en) | 2014-01-23 | 2014-01-23 | Method and device for classifying a behavior of a pedestrian when crossing a roadway of a vehicle and personal protection system of a vehicle |
| DE102014201159.0 | 2014-01-23 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104802793A true CN104802793A (en) | 2015-07-29 |
| CN104802793B CN104802793B (en) | 2019-07-26 |
Family
ID=53497872
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510030114.5A Active CN104802793B (en) | 2014-01-23 | 2015-01-21 | Passenger protection system for the method and apparatus and vehicle classified to behavior of the pedestrian in the runway for crossing vehicle |
Country Status (3)
| Country | Link |
|---|---|
| US (1) | US9734390B2 (en) |
| CN (1) | CN104802793B (en) |
| DE (1) | DE102014201159A1 (en) |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105151043A (en) * | 2015-08-19 | 2015-12-16 | 内蒙古麦酷智能车技术有限公司 | Emergency avoidance system and method for unmanned automobile |
| CN105911986A (en) * | 2016-04-25 | 2016-08-31 | 百度在线网络技术（北京）有限公司 | Unmanned vehicle perception test system and test method |
| CN106503627A (en) * | 2016-09-30 | 2017-03-15 | 西安翔迅科技有限责任公司 | A kind of vehicle based on video analysis avoids pedestrian detection method |
| CN106985781A (en) * | 2015-12-09 | 2017-07-28 | 罗伯特·博世有限公司 | The method and apparatus of device is alerted for triggering pedestrian protecting device and/or pedestrian |
| CN107792066A (en) * | 2016-09-01 | 2018-03-13 | 马自达汽车株式会社 | Controller of vehicle |
| CN108028019A (en) * | 2015-09-18 | 2018-05-11 | 大陆汽车有限责任公司 | Apparatus and method for vehicle startup auxiliary |
| CN108263277A (en) * | 2016-12-30 | 2018-07-10 | 现代自动车株式会社 | For alleviating the device and method of pedestrian impact |
| CN108725454A (en) * | 2017-04-25 | 2018-11-02 | 奥迪股份公司 | Safe driving assistant system and its control method |
| CN108733434A (en) * | 2018-05-21 | 2018-11-02 | 上海与德通讯技术有限公司 | The automatic avoiding method and Intelligent seat of seat |
| CN108803626A (en) * | 2018-08-16 | 2018-11-13 | 大连民族大学 | The system of Autonomous Vehicle or DAS (Driver Assistant System) programme path |
| CN108885835A (en) * | 2016-03-29 | 2018-11-23 | 松下知识产权经营株式会社 | Vehicle condition decision maker and vehicle condition determination method |
| CN108961838A (en) * | 2018-08-16 | 2018-12-07 | 大连民族大学 | Road pedestrian categorizing system |
| CN108974006A (en) * | 2017-06-01 | 2018-12-11 | 本田技研工业株式会社 | Prediction meanss, vehicle, prediction technique and storage medium |
| CN108985271A (en) * | 2018-08-16 | 2018-12-11 | 大连民族大学 | Pedestrian's determination method interested of magnetic force model |
| CN109063642A (en) * | 2018-08-01 | 2018-12-21 | 广州大学 | The prediction technique and system that pedestrian based on HMM algorithm goes across the road |
| CN109147389A (en) * | 2018-08-16 | 2019-01-04 | 大连民族大学 | The method of Autonomous Vehicle or DAS (Driver Assistant System) programme path |
| CN109147387A (en) * | 2017-06-19 | 2019-01-04 | 松下电器（美国）知识产权公司 | Information processing unit and program |
| CN109165591A (en) * | 2018-08-16 | 2019-01-08 | 大连民族大学 | road pedestrian classification method |
| CN109255970A (en) * | 2018-09-20 | 2019-01-22 | 南京锦和佳鑫信息科技有限公司 | A kind of intelligent network connection traffic safety system and method |
| CN109760674A (en) * | 2017-11-09 | 2019-05-17 | 丰田自动车株式会社 | Vehicle control apparatus |
| WO2019076542A3 (en) * | 2017-10-16 | 2019-06-13 | Robert Bosch Gmbh | Method and system for operating a vehicle |
| CN109969172A (en) * | 2017-12-26 | 2019-07-05 | 华为技术有限公司 | Control method for vehicle, equipment and computer storage medium |
| CN110758399A (en) * | 2018-07-27 | 2020-02-07 | 通用汽车环球科技运作有限责任公司 | System and method for predicting entity behavior |
| CN111094096A (en) * | 2017-09-29 | 2020-05-01 | 本田技研工业株式会社 | Vehicle control device, vehicle control method, and program |
| CN111243337A (en) * | 2020-02-09 | 2020-06-05 | 苏州商信宝信息科技有限公司 | Intelligent pedestrian protection method and system based on big data processing and analysis |
| CN112106348A (en) * | 2018-05-11 | 2020-12-18 | 阿诺·查斯 | Passive infrared pedestrian detection and avoidance system |
Families Citing this family (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9786177B2 (en) * | 2015-04-10 | 2017-10-10 | Honda Motor Co., Ltd. | Pedestrian path predictions |
| DE102015216881A1 (en) * | 2015-09-03 | 2017-03-09 | Robert Bosch Gmbh | Method and device for driverless driving a motor vehicle within a parking lot |
| US20180260635A1 (en) | 2015-09-08 | 2018-09-13 | Apple Inc. | Intention recognition |
| US10940856B2 (en) | 2016-01-18 | 2021-03-09 | Mitsubishi Electric Corporation | Driving assistance system |
| DE102017103700A1 (en) | 2017-02-23 | 2018-08-23 | Valeo Schalter Und Sensoren Gmbh | Avoiding a collision by a steering intervention in response to a movement of an external vehicle object |
| DE102017204404B3 (en) | 2017-03-16 | 2018-06-28 | Audi Ag | A method and predicting device for predicting a behavior of an object in an environment of a motor vehicle and a motor vehicle |
| DE102017010513A1 (en) | 2017-11-13 | 2018-05-09 | Daimler Ag | A method for detecting an attention of a person detected in an environment of a vehicle |
| US11206375B2 (en) | 2018-03-28 | 2021-12-21 | Gal Zuckerman | Analyzing past events by utilizing imagery data captured by a plurality of on-road vehicles |
| JP7024577B2 (en) * | 2018-04-24 | 2022-02-24 | 株式会社デンソー | Driving scene learning device, driving scene estimation device, and driving scene estimation system |
| US11138418B2 (en) | 2018-08-06 | 2021-10-05 | Gal Zuckerman | Systems and methods for tracking persons by utilizing imagery data captured by on-road vehicles |
| DE102018220690A1 (en) | 2018-11-30 | 2020-06-04 | Robert Bosch Gmbh | Method for classifying a behavior of a road user |
| JP7115277B2 (en) * | 2018-12-10 | 2022-08-09 | トヨタ自動車株式会社 | Behavior monitoring device, behavior monitoring system, and behavior monitoring program |
| DE102018133157A1 (en) | 2018-12-20 | 2020-06-25 | Daimler Ag | Method for the automated control of a vehicle at a pedestrian crossing, control device |
| WO2020127689A1 (en) | 2018-12-20 | 2020-06-25 | Robert Bosch Gmbh | Method for the automated control of a vehicle at a pedestrian crosswalk, control unit |
| KR20200139443A (en) * | 2019-06-04 | 2020-12-14 | 주식회사 만도 | Apparatus and method for driver assistance |
| KR20210053385A (en) * | 2019-11-01 | 2021-05-12 | 엘지전자 주식회사 | Vehicle Having a Dangerous Situation Notification Function and Control Method thereof |
| KR20210088780A (en) * | 2020-01-06 | 2021-07-15 | 주식회사 만도 | An apparatus for assisting driving of a vehicle and method thereof |
| CN113096424B (en) * | 2021-03-25 | 2022-05-31 | 武汉光庭信息技术股份有限公司 | Automatic emergency braking method and system for pedestrian crossing vehicle |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100076621A1 (en) * | 2007-04-02 | 2010-03-25 | Panasonic Corporation | Safety driving support apparatus |
| WO2010037823A1 (en) * | 2008-10-01 | 2010-04-08 | Universität Kassel | Method for avoiding collisions |
| JP2012117829A (en) * | 2010-11-29 | 2012-06-21 | Denso Corp | Prediction device |
| WO2013042205A1 (en) * | 2011-09-20 | 2013-03-28 | トヨタ自動車株式会社 | Pedestrian action prediction device and pedestrian action prediction method |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9302678B2 (en) * | 2006-12-29 | 2016-04-05 | Robotic Research, Llc | Robotic driving system |
| CN103814401B (en) * | 2011-09-20 | 2016-04-13 | 丰田自动车株式会社 | Object change detecting device and object change detecting method |
| US20130197736A1 (en) * | 2012-01-30 | 2013-08-01 | Google Inc. | Vehicle control based on perception uncertainty |
| JP5696701B2 (en) * | 2012-08-31 | 2015-04-08 | 株式会社デンソー | Anti-pedestrian notification device |
| WO2014162538A1 (en) * | 2013-04-03 | 2014-10-09 | トヨタ自動車株式会社 | Detection device, detection method, driving assistance device, and driving assistance method |
| JP6429368B2 (en) * | 2013-08-02 | 2018-11-28 | 本田技研工業株式会社 | Inter-vehicle communication system and method |
| JP2015072570A (en) * | 2013-10-02 | 2015-04-16 | 本田技研工業株式会社 | Vehicle controller |
-
2014
- 2014-01-23 DE DE102014201159.0A patent/DE102014201159A1/en active Pending
-
2015
- 2015-01-21 CN CN201510030114.5A patent/CN104802793B/en active Active
- 2015-01-23 US US14/604,130 patent/US9734390B2/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100076621A1 (en) * | 2007-04-02 | 2010-03-25 | Panasonic Corporation | Safety driving support apparatus |
| WO2010037823A1 (en) * | 2008-10-01 | 2010-04-08 | Universität Kassel | Method for avoiding collisions |
| JP2012117829A (en) * | 2010-11-29 | 2012-06-21 | Denso Corp | Prediction device |
| WO2013042205A1 (en) * | 2011-09-20 | 2013-03-28 | トヨタ自動車株式会社 | Pedestrian action prediction device and pedestrian action prediction method |
Cited By (36)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105151043A (en) * | 2015-08-19 | 2015-12-16 | 内蒙古麦酷智能车技术有限公司 | Emergency avoidance system and method for unmanned automobile |
| CN105151043B (en) * | 2015-08-19 | 2018-07-06 | 内蒙古麦酷智能车技术有限公司 | A kind of method of pilotless automobile Emergency avoidance |
| CN108028019A (en) * | 2015-09-18 | 2018-05-11 | 大陆汽车有限责任公司 | Apparatus and method for vehicle startup auxiliary |
| CN106985781A (en) * | 2015-12-09 | 2017-07-28 | 罗伯特·博世有限公司 | The method and apparatus of device is alerted for triggering pedestrian protecting device and/or pedestrian |
| CN108885835A (en) * | 2016-03-29 | 2018-11-23 | 松下知识产权经营株式会社 | Vehicle condition decision maker and vehicle condition determination method |
| CN105911986A (en) * | 2016-04-25 | 2016-08-31 | 百度在线网络技术（北京）有限公司 | Unmanned vehicle perception test system and test method |
| CN107792066A (en) * | 2016-09-01 | 2018-03-13 | 马自达汽车株式会社 | Controller of vehicle |
| CN106503627A (en) * | 2016-09-30 | 2017-03-15 | 西安翔迅科技有限责任公司 | A kind of vehicle based on video analysis avoids pedestrian detection method |
| CN108263277A (en) * | 2016-12-30 | 2018-07-10 | 现代自动车株式会社 | For alleviating the device and method of pedestrian impact |
| CN108725454B (en) * | 2017-04-25 | 2021-08-24 | 奥迪股份公司 | Safe driving assistance system and control method thereof |
| CN108725454A (en) * | 2017-04-25 | 2018-11-02 | 奥迪股份公司 | Safe driving assistant system and its control method |
| CN108974006B (en) * | 2017-06-01 | 2021-08-10 | 本田技研工业株式会社 | Prediction device |
| CN108974006A (en) * | 2017-06-01 | 2018-12-11 | 本田技研工业株式会社 | Prediction meanss, vehicle, prediction technique and storage medium |
| CN109147387A (en) * | 2017-06-19 | 2019-01-04 | 松下电器（美国）知识产权公司 | Information processing unit and program |
| CN111094096A (en) * | 2017-09-29 | 2020-05-01 | 本田技研工业株式会社 | Vehicle control device, vehicle control method, and program |
| US11498551B2 (en) | 2017-10-16 | 2022-11-15 | Robert Bosch Gmbh | Method and system for operating a vehicle |
| WO2019076542A3 (en) * | 2017-10-16 | 2019-06-13 | Robert Bosch Gmbh | Method and system for operating a vehicle |
| CN109760674B (en) * | 2017-11-09 | 2022-03-04 | 丰田自动车株式会社 | Vehicle control apparatus |
| CN109760674A (en) * | 2017-11-09 | 2019-05-17 | 丰田自动车株式会社 | Vehicle control apparatus |
| CN109969172B (en) * | 2017-12-26 | 2020-12-01 | 华为技术有限公司 | Vehicle control method, apparatus, and computer storage medium |
| CN109969172A (en) * | 2017-12-26 | 2019-07-05 | 华为技术有限公司 | Control method for vehicle, equipment and computer storage medium |
| CN112106348B (en) * | 2018-05-11 | 2024-03-15 | 阿诺·查斯 | Passive infrared pedestrian detection and avoidance system |
| CN112106348A (en) * | 2018-05-11 | 2020-12-18 | 阿诺·查斯 | Passive infrared pedestrian detection and avoidance system |
| CN108733434A (en) * | 2018-05-21 | 2018-11-02 | 上海与德通讯技术有限公司 | The automatic avoiding method and Intelligent seat of seat |
| CN110758399A (en) * | 2018-07-27 | 2020-02-07 | 通用汽车环球科技运作有限责任公司 | System and method for predicting entity behavior |
| CN110758399B (en) * | 2018-07-27 | 2023-04-18 | 通用汽车环球科技运作有限责任公司 | System and method for predicting entity behavior |
| CN109063642A (en) * | 2018-08-01 | 2018-12-21 | 广州大学 | The prediction technique and system that pedestrian based on HMM algorithm goes across the road |
| CN109165591A (en) * | 2018-08-16 | 2019-01-08 | 大连民族大学 | road pedestrian classification method |
| CN108803626B (en) * | 2018-08-16 | 2021-01-26 | 大连民族大学 | System for planning a route for an autonomous vehicle or a driver assistance system |
| CN109147389A (en) * | 2018-08-16 | 2019-01-04 | 大连民族大学 | The method of Autonomous Vehicle or DAS (Driver Assistant System) programme path |
| CN108985271B (en) * | 2018-08-16 | 2021-10-08 | 大连民族大学 | Method for judging pedestrian of interest of magnetic model |
| CN108985271A (en) * | 2018-08-16 | 2018-12-11 | 大连民族大学 | Pedestrian's determination method interested of magnetic force model |
| CN108961838A (en) * | 2018-08-16 | 2018-12-07 | 大连民族大学 | Road pedestrian categorizing system |
| CN108803626A (en) * | 2018-08-16 | 2018-11-13 | 大连民族大学 | The system of Autonomous Vehicle or DAS (Driver Assistant System) programme path |
| CN109255970A (en) * | 2018-09-20 | 2019-01-22 | 南京锦和佳鑫信息科技有限公司 | A kind of intelligent network connection traffic safety system and method |
| CN111243337A (en) * | 2020-02-09 | 2020-06-05 | 苏州商信宝信息科技有限公司 | Intelligent pedestrian protection method and system based on big data processing and analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104802793B (en) | 2019-07-26 |
| US9734390B2 (en) | 2017-08-15 |
| DE102014201159A1 (en) | 2015-07-23 |
| US20150206001A1 (en) | 2015-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104802793A (en) | Method and device for classifying a behavior of a pedestrian when crossing a roadway of a vehicle as well as passenger protection system of a vehicle | |
| CN113165652B (en) | Verifying predicted trajectories using a mesh-based approach | |
| US10332292B1 (en) | Vision augmentation for supplementing a person's view | |
| US11120691B2 (en) | Systems and methods for providing warnings to surrounding vehicles to avoid collisions | |
| JP4309843B2 (en) | Method and apparatus for preventing vehicle collision | |
| US11688174B2 (en) | System and method for determining vehicle data set familiarity | |
| CN103158705B (en) | Method and system for controlling a host vehicle | |
| KR20190058520A (en) | Location-based support for autonomous vehicle control systems | |
| US20180370502A1 (en) | Method and system for autonomous emergency self-learning braking for a vehicle | |
| JP7040374B2 (en) | Object detection device, vehicle control system, object detection method and computer program for object detection | |
| US9360332B2 (en) | Method for determining a course of a traffic lane for a vehicle | |
| US20130058116A1 (en) | Method and device for changing a light emission of at least one headlight of a vehicle | |
| US11351993B2 (en) | Systems and methods for adapting a driving assistance system according to the presence of a trailer | |
| CN110036426B (en) | Control device and control method | |
| JP6406356B2 (en) | Travel control device | |
| CN105599610A (en) | Road sign information display system and method | |
| US10803307B2 (en) | Vehicle control apparatus, vehicle, vehicle control method, and storage medium | |
| JP7289760B2 (en) | electronic controller | |
| US20200117950A1 (en) | System and method for evaluating a trained vehicle data set familiarity of a driver assitance system | |
| US11891058B2 (en) | Method and control unit for detecting a lane boundary | |
| CN113525386A (en) | Detection of rear approaching emergency vehicle | |
| US11430334B2 (en) | Systems and methods for adaptable rear-end collision alerts | |
| US11409297B2 (en) | Method for operating an automated vehicle | |
| Klette et al. | Vision-based driver assistance systems | |
| CN114375407A (en) | Method for operating a steering assistance system, steering assistance system and motor vehicle having such a steering assistance system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information |
Inventor after: T*Maurer Inventor after: T. Gessner Inventor after: Buerkle Lutz Inventor after: D - M - Gafulila Inventor before: T*Maurer Inventor before: T. Gessner Inventor before: Buerkle Lutz |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |